Generating One-Hot Maps Under Encryption

Ehud Aharoni; Nir Drucker; Eyal Kushnir; Ramy Masalha; Hayim Shaul

doi:10.1007/978-3-031-34671-2_8

Generating One-Hot Maps Under Encryption

Ehud Aharoni, Nir Drucker, Eyal Kushnir, Ramy Masalha, Hayim Shaul

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

Abstract

One-hot maps are commonly used in the AI domain. Unsurprisingly, they can also bring great benefits to ML-based algorithms such as decision trees that run under Homomorphic Encryption (HE), specifically CKKS. Prior studies in this domain used these maps but assumed that the client encrypts them. Here, we consider different tradeoffs that may affect the client’s decision on how to pack and store these maps. We suggest several conversion algorithms when working with encrypted data and report their costs. Our goal is to equip the ML over HE designer with the data it needs for implementing encrypted one-hot maps.

Original language	English
Title of host publication	Cyber Security, Cryptology, and Machine Learning - 7th International Symposium, CSCML 2023, Proceedings
Editors	Shlomi Dolev, Ehud Gudes, Pascal Paillier
Publisher	Springer Science and Business Media Deutschland GmbH
Pages	96-116
Number of pages	21
ISBN (Print)	9783031346705
DOIs	https://doi.org/10.1007/978-3-031-34671-2_8
State	Published - 2023
Externally published	Yes
Event	7th International Symposium on Cyber Security, Cryptology, and Machine Learning, CSCML 2023 - Be'er Sheva, Israel Duration: 29 Jun 2023 → 30 Jun 2023

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	13914 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	7th International Symposium on Cyber Security, Cryptology, and Machine Learning, CSCML 2023
Country/Territory	Israel
City	Be'er Sheva
Period	29/06/23 → 30/06/23

Bibliographical note

Publisher Copyright:
© 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.

Keywords

decision trees
homomorphic encryption
one-hot maps
PPML
privacy preserving transformation
privacy preservingmachine learning

ASJC Scopus subject areas

Theoretical Computer Science
General Computer Science

Access to Document

10.1007/978-3-031-34671-2_8

Cite this

Aharoni, E., Drucker, N., Kushnir, E., Masalha, R., & Shaul, H. (2023). Generating One-Hot Maps Under Encryption. In S. Dolev, E. Gudes, & P. Paillier (Eds.), Cyber Security, Cryptology, and Machine Learning - 7th International Symposium, CSCML 2023, Proceedings (pp. 96-116). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 13914 LNCS). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-3-031-34671-2_8

Aharoni, Ehud ; Drucker, Nir ; Kushnir, Eyal et al. / Generating One-Hot Maps Under Encryption. Cyber Security, Cryptology, and Machine Learning - 7th International Symposium, CSCML 2023, Proceedings. editor / Shlomi Dolev ; Ehud Gudes ; Pascal Paillier. Springer Science and Business Media Deutschland GmbH, 2023. pp. 96-116 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{30eea40a9a2f426498e5681cae23d07d,

title = "Generating One-Hot Maps Under Encryption",

abstract = "One-hot maps are commonly used in the AI domain. Unsurprisingly, they can also bring great benefits to ML-based algorithms such as decision trees that run under Homomorphic Encryption (HE), specifically CKKS. Prior studies in this domain used these maps but assumed that the client encrypts them. Here, we consider different tradeoffs that may affect the client{\textquoteright}s decision on how to pack and store these maps. We suggest several conversion algorithms when working with encrypted data and report their costs. Our goal is to equip the ML over HE designer with the data it needs for implementing encrypted one-hot maps.",

keywords = "decision trees, homomorphic encryption, one-hot maps, PPML, privacy preserving transformation, privacy preservingmachine learning",

author = "Ehud Aharoni and Nir Drucker and Eyal Kushnir and Ramy Masalha and Hayim Shaul",

note = "Publisher Copyright: {\textcopyright} 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.; 7th International Symposium on Cyber Security, Cryptology, and Machine Learning, CSCML 2023 ; Conference date: 29-06-2023 Through 30-06-2023",

year = "2023",

doi = "10.1007/978-3-031-34671-2_8",

language = "English",

isbn = "9783031346705",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Science and Business Media Deutschland GmbH",

pages = "96--116",

editor = "Shlomi Dolev and Ehud Gudes and Pascal Paillier",

booktitle = "Cyber Security, Cryptology, and Machine Learning - 7th International Symposium, CSCML 2023, Proceedings",

address = "Germany",

}

Aharoni, E, Drucker, N, Kushnir, E, Masalha, R & Shaul, H 2023, Generating One-Hot Maps Under Encryption. in S Dolev, E Gudes & P Paillier (eds), Cyber Security, Cryptology, and Machine Learning - 7th International Symposium, CSCML 2023, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 13914 LNCS, Springer Science and Business Media Deutschland GmbH, pp. 96-116, 7th International Symposium on Cyber Security, Cryptology, and Machine Learning, CSCML 2023, Be'er Sheva, Israel, 29/06/23. https://doi.org/10.1007/978-3-031-34671-2_8

Generating One-Hot Maps Under Encryption. / Aharoni, Ehud; Drucker, Nir; Kushnir, Eyal et al.
Cyber Security, Cryptology, and Machine Learning - 7th International Symposium, CSCML 2023, Proceedings. ed. / Shlomi Dolev; Ehud Gudes; Pascal Paillier. Springer Science and Business Media Deutschland GmbH, 2023. p. 96-116 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 13914 LNCS).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Generating One-Hot Maps Under Encryption

AU - Aharoni, Ehud

AU - Drucker, Nir

AU - Kushnir, Eyal

AU - Masalha, Ramy

AU - Shaul, Hayim

PY - 2023

Y1 - 2023

N2 - One-hot maps are commonly used in the AI domain. Unsurprisingly, they can also bring great benefits to ML-based algorithms such as decision trees that run under Homomorphic Encryption (HE), specifically CKKS. Prior studies in this domain used these maps but assumed that the client encrypts them. Here, we consider different tradeoffs that may affect the client’s decision on how to pack and store these maps. We suggest several conversion algorithms when working with encrypted data and report their costs. Our goal is to equip the ML over HE designer with the data it needs for implementing encrypted one-hot maps.

AB - One-hot maps are commonly used in the AI domain. Unsurprisingly, they can also bring great benefits to ML-based algorithms such as decision trees that run under Homomorphic Encryption (HE), specifically CKKS. Prior studies in this domain used these maps but assumed that the client encrypts them. Here, we consider different tradeoffs that may affect the client’s decision on how to pack and store these maps. We suggest several conversion algorithms when working with encrypted data and report their costs. Our goal is to equip the ML over HE designer with the data it needs for implementing encrypted one-hot maps.

KW - decision trees

KW - homomorphic encryption

KW - one-hot maps

KW - PPML

KW - privacy preserving transformation

KW - privacy preservingmachine learning

UR - http://www.scopus.com/inward/record.url?scp=85164969698&partnerID=8YFLogxK

U2 - 10.1007/978-3-031-34671-2_8

DO - 10.1007/978-3-031-34671-2_8

M3 - Conference contribution

AN - SCOPUS:85164969698

SN - 9783031346705

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 96

EP - 116

BT - Cyber Security, Cryptology, and Machine Learning - 7th International Symposium, CSCML 2023, Proceedings

A2 - Dolev, Shlomi

A2 - Gudes, Ehud

A2 - Paillier, Pascal

PB - Springer Science and Business Media Deutschland GmbH

T2 - 7th International Symposium on Cyber Security, Cryptology, and Machine Learning, CSCML 2023

Y2 - 29 June 2023 through 30 June 2023

ER -

Aharoni E, Drucker N, Kushnir E, Masalha R, Shaul H. Generating One-Hot Maps Under Encryption. In Dolev S, Gudes E, Paillier P, editors, Cyber Security, Cryptology, and Machine Learning - 7th International Symposium, CSCML 2023, Proceedings. Springer Science and Business Media Deutschland GmbH. 2023. p. 96-116. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-031-34671-2_8

Generating One-Hot Maps Under Encryption

Abstract

Publication series

Conference

Bibliographical note

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this